(LRV1) Pathogenicity Factor

Home , Cutaneous leishmaniasis, Leishmaniavirus, Leishmania aethiopica, Leishmania braziliensis

Antiviral screening identifies adenosine analogs PNAS PLUS targeting the endogenous dsRNA Leishmania RNA virus 1 (LRV1) pathogenicity factor

F. Matthew Kuhlmanna,b, John I. Robinsona, Gregory R. Bluemlingc, Catherine Ronetd, Nicolas Faseld, and Stephen M. Beverleya,1

aDepartment of Molecular Microbiology, Washington University School of Medicine in St. Louis, St. Louis, MO 63110; bDepartment of Medicine, Division of Infectious Diseases, Washington University School of Medicine in St. Louis, St. Louis, MO 63110; cEmory Institute for Drug Development, Emory University, Atlanta, GA 30329; and dDepartment of Biochemistry, University of Lausanne, 1066 Lausanne, Switzerland

Contributed by Stephen M. Beverley, December 19, 2016 (sent for review November 21, 2016; reviewed by Buddy Ullman and C. C. Wang) + + The endogenous double-stranded RNA (dsRNA) virus Leishmaniavirus macrophages infected in vitro with LRV1 L. guyanensis or LRV2 (LRV1) has been implicated as a pathogenicity factor for leishmaniasis Leishmania aethiopica release higher levels of cytokines, which are in rodent models and human disease, and associated with drug-treat- dependent on Toll-like receptor 3 (7, 10). Recently, methods for ment failures in Leishmania braziliensis and Leishmania guyanensis systematically eliminating LRV1 by RNA interference have been − infections. Thus, methods targeting LRV1 could have therapeutic ben- developed, enabling the generation of isogenic LRV1 lines and efit. Here we screened a panel of antivirals for parasite and LRV1 allowing the extension of the LRV1-dependent virulence paradigm inhibition, focusing on nucleoside analogs to capitalize on the highly to L. braziliensis (12). active salvage pathways of Leishmania, which are purine auxo- A key question is the relevancy of the studies carried out in trophs. Applying a capsid flow cytometry assay, we identified two murine models to human disease. For L. guyanensis, patients + 2′-C-methyladenosine analogs showing selective inhibition of LRV1. infected with LRV1 strains show an increased severity of cu- Treatment resulted in loss of LRV1 with first-order kinetics, as taneous disease (13). In humans, L. braziliensis is associated with expected for random virus segregation, and elimination within six cutaneous leishmaniasis, as well as the larger share of the more MICROBIOLOGY cell doublings, consistent with a measured LRV1 copy number of debilitating mucocutaneous leishmaniasis (MCL). Thus far there about 15. Viral loss was specific to antiviral nucleoside treatment are no data available in humans permitting tests of the associa- and not induced by growth inhibitors, in contrast to fungal dsRNA + − tion of LRV1 with L. braziliensis parasite burden nor the severity viruses. Comparisons of drug-treated LRV1 and LRV1 lines recapit- of cutaneous leishmaniasis (CL), which can show a range of ulated LRV1-dependent pathology and parasite replication in mouse presentations (14, 15). In lieu of such information, studies have infections, and cytokine secretion in macrophage infections. Agents focused on the association of LRV1 with MCL vs. CL, which is targeting Totiviridae have not been described previously, nor are thought to reflect primarily immunopathology rather than par- there many examples of inhibitors acting against dsRNA viruses more – generally. The compounds identified here provide a key proof-of- asite numbers (2, 6, 14 16). Although in some studies LRV1 was principle in support of further studies identifying efficacious antivi- not correlated with MCL clinical manifestations (17, 18), in rals for use in in vivo studies of LRV1-mediated virulence. others there was a strong association (6, 19, 20). The basis for these discrepancies is of considerable interest, hypotheses for trypanosomatid protozoan parasite | endobiont virus | viral segregation | which include other parasite or host factors known to play a chemotherapy | virulence significant role in the development of MCL (13, 21, 22), or

rotozoan parasites of the genus Leishmania are responsible Significance Pfor leishmaniasis in many regions of the world, with 12 million current cases (accompanied by at least 10-fold more bearing The endogenous double-stranded RNA virus Leishmaniavirus asymptomatic infections) and nearly 1.7 billion people at risk (1–5). (LRV1) has been implicated as a pathogenicity factor for leish- The disease has three predominant clinical manifestations, ranging maniasis in rodent models and human disease, and associated from the relatively mild, self-healing cutaneous form, to mucocuta- with drug-treatment failures. As a first step toward the identifi- neous lesions where parasites metastasize to and cause destruction cation of therapeutic LRV1 inhibitors, we identified two adeno- of mucous membranes of the nose, mouth, and throat, or fatal sine analogs that selectively inhibited LRV1 replication. These visceral disease. Disease phenotypes segregate primarily with analogs were used as tools to confirm that viral inheritance is by the infecting species; however, it is not fully understood which random segregation, as well as to generate LRV1-cured lines of parasite factors affect severity and disease manifestations. Leishmania guyanensis, which correspondingly lost the increased One recently identified parasite factor contributing to disease se- pathogenicity conferred by LRV1. These compounds hold promise verity in several Leishmania speciesistheRNAvirusLeishmaniavirus as leads to ameliorate the severity of LRV1-bearing Leishmania (6, 7). These endobiont viruses classified within the Totiviridae are infections, and raise the possibility of targeting other protozoal comprised of a single-segmented double-stranded RNA (dsRNA) infections whose pathogenicity may be exacerbated by similar genome that encodes only a capsid protein and an RNA-dependent endogenous viruses. RNA polymerase (RDRP) (8, 9). Leishmaniavirus is most frequently Author contributions: F.M.K., N.F., and S.M.B. designed research; F.M.K., J.I.R., and C.R. found in New World parasite species in the subgenus Viannia [as performed research; G.R.B. contributed new reagents/analytic tools; F.M.K., J.I.R., C.R., Leishmania RNA virus 1 (LRV1)], such as Leishmania braziliensis and S.M.B. analyzed data; and F.M.K. and S.M.B. wrote the paper. (Lbr)andLeishmania guyanensis (Lgy), which cause both cutaneous Reviewers: B.U., Oregon Health Sciences University; and C.C.W., University of California, and mucocutaneous disease (6), and is found sporadically in Old San Francisco. World subgenus Leishmania species [as Leishmania RNA virus 2 The authors declare no conflict of interest. (LRV2)] (10, 11). Mice infected with LRV1-bearing strains of 1To whom correspondence should be addressed. Email: [email protected]. L. guyanensis exhibit greater footpad swelling and higher parasite − This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. numbersthanmiceinfectedwithLRV1 L. guyanensis (7). Similarly, 1073/pnas.1619114114/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1619114114 PNAS Early Edition | 1of9 Downloaded by guest on September 25, 2021 microbial sources including the microbiota or coinfections (23). Recent studies show that the presence of LRV1 in clinical isolates of L. braziliensis and L. guyanensis correlates with drug-treatment failure (17, 20), phenomena that could readily be explained by the increased parasite numbers or altered host responses predicted from animal models (7, 13, 24). Thus, current data support a role for LRV1 in increasing disease severity in human leishmaniasis (13); this suggests that therapies targeting LRV1 specifically could be applied toward amelioration of disease pathology. As one approach, murine vaccination using the LRV1 capsid results in significant + protection against LRV1 L. guyanensis (25). Here we describe a complementary approach, targeting LRV1 directly using small-molecule inhibitors. Although effective antivirals are available for many viral targets including retroviruses, DNA viruses, and single-strand RNA (ssRNA) viruses (26), little effort has gone into agents acting against dsRNA viruses (27). These comprise at least 10 viral families (Birnaviridae, Boty- birndaviridae, Chrysoviridae, Cystoviridae, Megabirnavirdae, Partitiviridae, Picobirnaviridae, Quadriviridae, Reoviridae, and Totiviridae), infecting a wide array of hosts, including fungi, plants, and animals (28). Some constitute important agricultural pathogens and rotaviruses (Reoviridae) cause serious human disease. For protozoan viruses, their role in the exacerbation of human disease is only now beginning to be appreciated (6, 29). Because viral ele- ments are critical factors acting to exacerbate the disease where studied, candidate anti-LRV1 agents should be viewed as “anti- pathogenicity” treatments rather than sterilizing cures (30), which could be used alone or more likely in combination with existing antileishmanial agents in the treatment of ongoing infection. As a starting point, we focused on nucleoside analogs, a class that includes many widely used and effective antivirals (Table S1) (26). Following uptake and activation to the triphosphate form, these analogs primarily target viral replication, with different classes acting preferentially against viral DNA or RNA + − polymerases (RDRP) or reverse transcriptases, as well as cellular Fig. 1. Anti-LRV1 capsid flow cytometry. LgyLRV1 and LgyLRV1 parasites metabolism. A second rationale was that Leishmania are purine were fixed and permeabilized followed by staining with increasing dilutions of auxotrophs, with highly active and multiply redundant pathways anticapsid antibody and fluoresceinated secondary antibody. (A) Profiles + − for uptake and activation of nucleobases and nucleosides (31). obtained with LgyLRV1 (solid line) and LgyLRV1 (filled) after selection for Indeed, a great deal of prior effort has been devoted to the de- single cells. A representative experiment is shown, performed at a dilution of velopment of antileishmanial purine analogs; however, whereas 1:16,000; subsequent studies were performed using a dilution of 1:20,000 (n > 11). (B) Mean fluorescence of LgyLRV1+ (■)andLgyLRV1− (□) for each the nucleobase allopurinol is commonly used as a veterinary agent, + − ● it has proven more difficult to find agents of sufficient potency and antibody dilution. The ratio of LRV1 /LRV1 staining ( ) is plotted as a solid line. selectivity against Leishmania to be used widely against human leishmaniasis (32). We reasoned that the highly divergent prop- to identify fixation conditions that gave similarly clear discrimi- + − erties of Totiviridae RDRPs, relative to the polymerases of both nation between LgyLRV1 and LRV1 by flow cytometry. the Leishmania and mammalian hosts (as well as other viral RDRPs), could prove fertile grounds for antiviral discovery, es- Inhibition Tests. We acquired a collection of 81 compounds, pri- pecially when coupled with potentiation by the parasite’s powerful marily nucleoside or nucleobase analogs, including ones shown nucleoside/base salvage pathways. previously to be active against diverse viruses, tumor cells, or Leishmania (Fig. S1 and Tables S1 and S2). These compounds + Results were examined for their ability to inhibit the growth of LgyLRV1 − Measurement of LRV1 Levels by Capsid Flow Cytometry. Because and virus levels by LRV1 capsid flow cytometry. LgyLRV1 par- + LRV1 (like most Totiviridae) is not shed from the cell (33, 34), asites grew similarly to LgyLRV1 and were used as virus-negative we developed a flow cytometric assay to measure intracellular controls. These data revealed three patterns (Fig. 2). For most LRV1 capsid levels on a per cell basis. To detect LRV1 we used compounds, LRV1 capsid levels were not significantly affected, binding to a rabbit anti-LgyLRV1 capsid antiserum (35) followed within a factor of ∼3 (Fig. 2, black or red dots within large dashed by detection with Alexa Fluor488-conjugated goat anti-rabbit gray and red circles, Fig. S2,andTable S2). All nucleobase analogs IgG. We found that fixation with 2% (wt/vol) paraformaldehyde fell within this group, as did foscarnet (a structure analog of py- followed by permeabilization with Triton X-100 yielded a clear rophosphate). Within this group, a subset showed more than LRV1-dependent profile (Fig. 1A). Titration of the anticapsid 10-fold inhibition of L. guyanensis growth (Fig. 2, red dashed circle antiserum showed that dilutions around 1:16,000 gave a strong and black dots above; Fig. S3 A and B;andTable S2), including + − signal with excellent selectivity between LgyLRV1 and LRV1 known antileishmanials, such as allopurinol, mycophenolic acid, (Fig. 1B), with little background staining evident in immunofluo- and 4-aminopyrazolopyrimidine (APP). Several additional com- rescence microscopy. Under these conditions and as seen in pre- pounds showed leishmanial inhibition at the concentration tested + vious immunofluorescence studies (36), LgyLRV1 showed a (Fig. 2, Fig. S2B,andTable S2); however, these were deprioritized strong, homogeneous LRV1 distribution (Fig. 1A). We attempted for various reasons, including known mammalian cell toxicity. In similar studies with anti-dsRNA antibodies (36), but were unable the initial screens several compounds showed modest elevation of

2of9 | www.pnas.org/cgi/doi/10.1073/pnas.1619114114 Kuhlmann et al. Downloaded by guest on September 25, 2021 PNAS PLUS (36), which showed an EC50 of ∼1 μM, slightly less than seen with capsid inhibition and consistent with the anticipated targeting of the RDRP. With 7d2CMA, an EC50 of ∼5 μM was seen against LRV1 capsid expression, versus ∼ >100 μMfor L. guyanensis growth, again with about >20-fold selectivity (Fig. 3B). Several studies were carried out with L. braziliensis strains bearing LRV1 (12). The 2CMA EC50 for LbrLRV1 was similar to that seen with LgyLRV1 (∼3 μM); however, parasites were somewhat more susceptible to growth in inhibition (EC50 50–100 μM). Because the available quantities of 7d2CMA were limiting and both compounds were similarly selective for L. guyanensis, wefocusedthereafteron2CMA.

Inhibition of 2CMA LRV1 Is Unaffected by Exogenous Adenine, nor Is Synergy Seen with Antileishmanial Nucleobases. We asked whether the 2CMA potency was affected by the presence of exogenous adenine, present at about 5–33 μM in the yeast extract compo- nent of Schneider’s medium (43). The addition of adenine up to 400 μM had no impact on LRV1 inhibition by 100 μM 2CMA, Fig. 2. Antiviral inhibition of L. guyanensis growth vs. LRV1 inhibition. The + figure shows data from Table S2 plotted; LRV1 capsid levels (y axis) vs. nor did it alter LRV1 levels in LgyLRV1 (Fig. S3C). APP L. guyanensis growth (x axis). The large dashed gray circle marks compounds showed similar inhibition of L. guyanensis growth and LRV1 (black dots) showing little effect on LRV1 or L. guyanensis, the red circle levels, whereas at the highest concentration tested, allopurinol marks compounds preferentially inhibiting L. guyanensis growth, and the inhibited L. guyanensis growth or LRV1 capsid levels by 30 or green circle marks compounds preferentially inhibiting LRV1; blue dots de- + 60%, respectively (Fig. S3A). We then explored potential in- pict 2′C substituted nucleosides without anti-LRV1 activity. LgyLRV1 and − teractions between 2CMA and antileishmanial nucleobases. LgyLRV1 controls are shown in brown. Abbreviations for compounds dis- cussed further in the text can be found in Table S1. However, no change in the EC50 for 2CMA inhibition of L. guyanensis growth or LRV1 capsid synthesis was seen with

increasing concentrations of allopurinol (∼3 μM) (Fig. S3D). MICROBIOLOGY LRV1, often accompanied by growth inhibition (Fig. 2, Fig. S2A, and Table S2). However, this effect was not always reproducible and was not pursued further. Two compounds strongly reduced LRV1 capsid levels with minimal impact on parasite growth (Fig. 2, green circle, Fig. S1 and Table S2). Both 2′-C-methyladenosine (2CMA) and 7-deaza-2′-C-methyladenosine (7d2CMA) resulted in 12-fold re- ductions in LRV1 capsid levels, showing 30% and 90% inhibition of parasite density, respectively, when tested at 100 μM. Both had previously been shown to inhibit the hepatitis C virus (HCV) RDRP following activation (37, 38). In contrast, 2′-C-methylcytidine or guanosine had little effect on LRV1 levels or L. guyanensis growth (Fig. 2, blue dots). Compounds bearing a variety of other 2′ modifications (alone or in combination, with various bases) showed little effect on LRV1. These included sofosbuvir and mericitabine (related to 2′-C-methyl-2′-F uridine or cytidine, respectively), both of which show strong activity against HCV (39, 40), or NITD008, which shows good activity against flaviviruses (41). These data suggest a strong preference for both the nature of the 2′-C sub- stitution, as well as adenine as the base. Note that these data cannot discriminate between effects arising from direct inhibition of RDRP or other viral processes, nor drug metabolism (phosphorylation or resistance to nucleoside hydrolases). Previously, a Leishmania cysteine proteinase activity was implicated in the cleavage of the LRV1 capsid–RDRP fusion protein, potentially important for LRV1 biogenesis (42). How- ever, no effects on L. guyanensis growth and only minimal effects on LRV1 capsid levels were observed with three cysteine proteinase inhibitors tested (E64, E64d, and CA-074) (Table S2), relative to the effects of 2CMA or 7d2CMA.

2CMA Preferentially Inhibits LRV1 Replication. Titrations were performed to quantitate the potency of 2CMA and 7d2CMA against L. guyanensis growth and LRV1, measuring the relative cellular Fig. 3. 2CMA and 7d2CMA inhibition of L. guyanensis growth and LRV1 capsid growth rate to better assess fitness effects. For 2CMA, the EC50 ∼ μ orRNAlevels.ThefigureshowstherateofgrowthorLRV1capsidlevels(y axis) was estimated to be 3 M for LRV1 capsid inhibition, versus ● > μ as a function of drug concentration. (A)2CMA;(B) 7d2CMA. Growth rate ( , 100 M for parasite growth (Fig. 3A), at least 30-fold selective. solid line) and normalized LRV1 capsid (□, dashed line) or RNA (Δ, dashed line) To assess the effects on replication of the dsRNA LRV1 genome are shown. The results of one representative experiment are shown for 2CMA directly, we used quantitative anti-dsRNA slot blots (Fig. 3A) (n = 2 for RNA and capsid) and a single experiment for 7d2CMA.

Kuhlmann et al. PNAS Early Edition | 3of9 Downloaded by guest on September 25, 2021 + LRV1 Inhibition Is Independent of Leishmania Growth Inhibition. growing LgyLRV1 in 100 μM 2CMA for one, three, four, or six Agents inducing stress or growth arrest have been used to cure cell doublings followed by transfer to drug-free media. After one fungal Totiviridae, with cycloheximide (CHX) used often (44, 45). doubling, a time when LRV1 levels had only decreased twofold, Growth of L. guyanensis at 10 or 100 nM CHX resulted in an LRV1 capsid levels rapidly returned to WT levels and distribution. increase in population doubling time, from ∼7.7 h to 11.2 or In contrast, when 2CMA was maintained for three or four cell 44.7 h, respectively, without significant cell death as evidenced by doublings, resulting in a homogeneous population showing on resumption of WT growth following CHX removal (Fig. 4A). average 8- or 16-fold less LRV1 capsid expression, the washout Despite the strong effects on growth, LRV1 capsid levels were lines now showed two distinct populations (Fig. 5 C and D). One + “ −” population expressed LRV1 at levels similar to control LgyLRV1 , unaffected, nor was the emergence of a LRV1 parasite pop- − whereas the other resembled LgyLRV1 (Fig. 5D, Top and ulation seen at any CHX concentration (Figs. 4 B and C). Similar + results were obtained with clotrimazole, which inhibits Leishmania Middle). Parasites with LgyLRV1 capsid levels were the majority growth through inhibition of sterol synthesis (Fig. 4D). Finally, no (55%) in the three-doubling washout population, whereas these had declined to 36% percent in the four-doubling washout pop- correlation was seen between LRV1 levels and growth rate in our − test compound screening (Fig. 2 and Fig. S2) or exposure to ulation (Fig. 5D). The LgyLRV1 population increased from 31 to hygromycin B (46). Thus, inhibition of Leishmania growth alone 50% of the total cell population during this time. Finally, after six cell doublings of growth with 2CMA, the LRV1 capsid profile was does not alter LRV1 levels. − indistinguishable from that of the LgyLRV1 and the six-doubling washout population revealed only parasites maintaining the Viral Loss Occurs by Random Dilution. The availability of an in- − hibitor with strong selectivity for LRV1 over parasite growth LgyLRV1 capsid-staining profile (Fig. 5D). This population was maintained for at least six passages (∼40 cell doublings) provided an opportunity to test the assumption that cytosolic + Totiviruses are passed randomly to daughter cells during mitosis without return of any demonstrable LRV1 parasites. (34, 47). For maximal LRV1 inhibition, parasites were in- Several conclusions emerge from these studies: first, the ef- μ fective LRV1 copy number per cell must be relatively low, as oculated into 100 M 2CMA, which increased the population − doubling time from 6.4 to 8.5 h (Fig. 3). The average LRV1 otherwise an LRV1 population could not emerge after only levels immediately declined, with capsid and RNA levels falling three to six cell doublings (Fig. 5), roughly corresponding to copy – 3 − 6 − in parallel (Fig. 5 A and B). Importantly, when plotted as a numbers of 8 64 (2 2 ) and consistent with fraction of LRV1 cells emerging in the washouts (Fig. 5D). LgyLRV1 copy number function of number of cell divisions, loss of LRV1 capsid and – RNA followed a first-order linear relationship, with a 50% loss was previously estimated as 24 100 by competitive PCR assay (48). To assess LRV1 copy number independently in the clonal at every doubling (Fig. 5 A and B). When visualized at the + population level by flow cytometry, LRV1 capsid levels per cell LgyLRV1 line studied here, we isolated total RNA quantita- declined homogeneously at every time point tested until only tively from a known number of cells, and estimated LRV1 copy number by quantitative RT-PCR (qRT-PCR), using a standard background staining was evident by six cell doublings (Fig. 5C). curve established from a cloned LRV1 genome (Methods). This Both of these observations closely match the expectation for the process yielded an estimated average LRV1 copy number of random distribution of LRV1 particles to daughter cells during 15 ± 0.9 per cell (n = 3), consistent with range estimated from mitosis and successive cell divisions. the rate of drug-induced loss above. − Second, after washout, 2CMA-treated parasites, which origi- 2CMA Induces LRV1 Populations. To explore the loss of LRV1 further, we performed a series of “washout” experiments, nally showed homogeneous low levels of LRV1, now reverted to biphasic populations showing WT or “negative” LRV1 levels. The recovery of the WT-like population suggests that there may be a “set point” for LRV1 levels. Because only populations but not clones were studied, we cannot be sure that this occurred intracellularly; however, the rapidity with which LRV1 levels rebounded suggests this may be more likely.

Rapid Recovery of Matched Clonal WT and LRV1-Cured Lines. Our findings suggested that it should be relatively easy to recover − LRV1 clonal lines from the 2CMA-treated population. How- ever, we were concerned that despite small effects on growth, the relatively high concentration of 2CMA used above could itself have unwanted selective effects on L. guyanensis. Support for this concern arose when in pilot studies, several clonal lines obtained after growth in 100 μM 2CMA lacked LRV1 but showed decreased growth inhibition by 2CMA. Thus, we repeated the LRV1 cure using 10 μM 2CMA, a concentration showing less of an effect on parasite growth but retaining strong inhibition of LRV1 levels (Fig. 3). Again, loss of LRV1 proceeded homogeneously (Fig. 6A). When clonal lines were recovered directly by − plating from this population, very few were LRV1 (1 of 30). However, if the population was allowed to grow in the absence of Fig. 4. LRV1 levels are unaffected by agents inhibiting L. guyanensis ∼ + 2CMA for another six cell doublings (washout), a bimodal growth. (A) LgyLRV1 was treated with 10 nM CHX (□, dashed line), 100 nM population for LRV1 capsid levels emerged, as seen previously. ○ ■ CHX ( , dashed line), or no treatment ( , solid line). After 72 h, cells treated Analysis of 12 clonal lines obtained by direct plating from this with 100 nM CHX were placed into fresh media (●, dashed line). (B) Profiles washout population showed that six exhibited LRV1 capsid obtained by LRV1 flow cytometry after 48 h growth for WT (shaded) or cells − levels/profiles identical to the LgyLRV1 control, whereas two treated with 100 μM CHX (solid line), or 10 μM CHX (dashed line). (C) Plot of + + growth rate of LgyLRV1 (●) or LRV1 capsid levels (□, dashed line) after 48-h showed profiles identical to the LgyLRV1 parent (representa- propagation in increasing concentrations of CHX. (D)AsinC but for clo- tives shown in Fig. 6B). Interestingly, four lines showed more trimazole (CTZ). A representative experiment is shown (n = 3). complex profiles, with populations showing range of intensities

4of9 | www.pnas.org/cgi/doi/10.1073/pnas.1619114114 Kuhlmann et al. Downloaded by guest on September 25, 2021 PNAS PLUS

+ Fig. 5. Kinetics of and cellular distribution of LRV1 loss after treatment with 100 μM 2CMA. (A and B) LgyLRV1 was inoculated into media without (●)or with (□, Δ) 100 μM 2CMA, and growth and LRV1 capsid (□, dashed line) and RNA levels (Δ, dashed line) measured by capsid flow cytometry (A) or qRT-PCR (B). + − − + − For A, results at each time are shown normalized to LRV1 and LRV1 control levels using the formula log2 (2CMA treated − LRV1 )/(LRV1 − LRV1 ). For B, the log2 ddCT values are shown. A theoretical 1:2 dilutional loss is shown (thin gray line); error bars represent ±1SD.(C) LRV1 capsid flow cytometry of control parasites (Top) and populations grown for one, three, four, or six cell doublings in 100 μM 2CMA. (D) LRV1 capsid flow cytometry of parasites grown for three, four, or six doublings in 100 μM 2CMA, and then grown for an additional six cell doublings in drug-free media (washouts). Thick and thin gray dashed lines + − represent LgyLRV1 and LgyLRV1 , respectively. MICROBIOLOGY − + spanning those from LRV1 to LRV1 controls (representative Leishmaniavirus, basic studies of viruses within the Totiviridae, shown in Fig. 6B). These complex lines were not studied further. the development of antivirals directed against dsRNA viruses The set-point hypothesis predicts that upon further growth, those generally, and the development of new tools for assessing the role lines would ultimately revert to bimodal populations. of LRV1 in elevating Leishmania pathogenicity. + We chose two LRV1 and LRV1-cured lines that had expe- To facilitate the search for LRV1 inhibitors, we first de- rienced identical 2CMA treatment and culture manipulations. veloped a capsid flow cytometry assay to rapidly monitor LRV1 Growth tests confirmed these were not resistant to 2CMA, and capsid levels (Fig. 1). This assay can be performed in only a few RT-PCR and Western blot tests confirmed the presence or ab- hours, and although these studies used it in a relatively low sence of LRV1 (Fig. 6 C and D). These clones thus constituted throughput manner, it should be scalable for higher throughput. matched WT and LRV1-cured lines appropriate for subsequent The results were confirmed by anticapsid or anti-dsRNA West- studies of LRV1 effects. ern or slot blotting, or quantitative RT-PCR (Fig. 6 C and D). Additionally, this assay provides useful information about the LRV1 Correlates with Increased Cytokine Secretion and Mouse Infectivity. + − cellular heterogeneity of LRV1 levels not readily achievable by With matched 2CMA-treated LRV1 and LRV1 (cured) lines, we other methods, which informed studies probing the inheritance − asked whether LRV1 was correlated with elevated pathology and of LRV1 as well as in the generation of LRV1 lines. hyperinflammatory responses, as expected (7, 12). Infections were We focused on known antivirals for several reasons: first, de- performed with bone marrow-derived macrophages (BMM) in vitro, spite significant advances in targeting many retroviruses, DNA followed by assays for secretion of two characteristic LRV1-dependent viruses, or ssRNA viruses, very little effort or progress has been cytokine reporters, IL-6 and TNF-α. Cytokine secretion induced by + devoted on inhibition of dsRNA viruses. Thus, there seemed a the LRV1 /2CMA-treated lines was comparable to that of the pa- + reasonable potential for “repurposing” known antivirals against rental LgyLRV1 line, whereas cytokine secretion induced by the 2CMA cured lines was considerably less, and comparable to that of the dsRNA Leishmaniavirus. Moreover, because many antivirals − the LgyLRV1 control (Fig. 7 A and B). are nucleoside analogs and that Leishmania is a purine auxotroph Infections of susceptible BALB/c mice were performed fol- (31), the pharmacokinetics of drug uptake and metabolism could lowed by measurement of pathology (footpad swelling) and well favor the efficacy of such compounds against Leishmaniavirus. bioluminescent imaging of parasite numbers. A strong LRV1- As a collateral benefit, these studies had the potential to uncover dependency for both pathology and parasite abundance was new lead inhibitors against Leishmania itself, as auxotrophy has + observed in comparisons of the matched 2CMA-treated LRV1 prompted many investigators to target purine metabolism for − vs. LRV1 (cured) lines (Fig. 7 C and D). Importantly, the re- antileishmanial therapy. Several new compounds not previously + sponse to the 2CMA-treated LRV1 lines closely matches that reported to inhibit Leishmania were identified (Fig. 2 and Tables + to the control parental LgyLRV1 line and, similarly, the re- S1 and S2), but were not pursued further here. − sponse to the 2CMA-treated LRV1 line closely matches that to We identified two compounds that showed preferential inhibition − the LgyLRV1 control (Fig. 7 C and D), both of which were of LRV1, 2CMA, or 72CMA (Fig. 3 and Fig. S1). The two active studied previously (7). compounds were effective in the micromolar range, with >20-fold selectivity for LRV1 versus L. guyanensis growth inhibition and Discussion were also active against LbrLRV1, albeit with somewhat less se- In this study, we report the identification of compounds specif- lectivity over growth. The EC50 measured using dsRNA or capsid ically targeting the LRV1 dsRNA virus of L. guyanensis and levels were similar, with that of the dsRNA being somewhat less, L. braziliensis, two representatives of the Totiviridae. Our find- consistent with the anticipated mode of action targeting the RDRP ings have relevance for the specific therapeutic inhibition of and genome replication. Both compounds have demonstrated

Kuhlmann et al. PNAS Early Edition | 5of9 Downloaded by guest on September 25, 2021 Other factors may include differential ability to be phosphorylated, often the rate-limiting step for antiviral nucleoside activation (52, 53), or susceptibility to nucleoside hydrolases or phosphorylases, which Leishmania possess in abundance (31), and affinity of the phosphorylated analog with the LRV1 RDRP itself. Additional studies will be required to assess the contributions of each of these factors to anti-LRV1 activity and the design of more potent inhibitors.

Anti-LRV1 Agents as a Tool for Studying Leishmaniavirus Replication and Biology. The LRV1 selectivity of 2CMA and 7d2CMA provided the foundation for several studies probing LRV1 biology. Under 2CMA inhibition, a first-order kinetic loss of LRV1 was observed, (measured by either capsid or dsRNA genome levels), with a homogeneous 50% loss at every cell doubling (Fig. 5 A and B). This finding fits exactly the prediction assumed by a random-inheritance model of LRV1 particles during mitosis. Although widely assumed for the inheritance of most persistent dsRNA viral infections, these findings now provide direct evidence of random segregation. These data also provide a mech- anistic explanation for the failure to identify compounds inhibiting both LRV1 and L. guyanensis in our screen, because without continued parasite growth LRV1 cannot be lost by dilution, and indeed may increase somewhat (Fig. 2). Ultimately, LRV1 levels declined to levels approaching those of LRV-free parasites within three to six cell doublings following 2CMA treatment (Fig. 5). This finding implies the viral copy – number was relatively low, less than 8–64 (23 6), significantly less than previous estimates of 120 for LgyLRV1 and often many + Fig. 6. Generation of matched LRV1 and cured lines after limited 10 μM thousands for other Totiviridae (34, 48). However, quantitative 2CMA treatment. (A) Workflow for treatment of parasites with 10 μM 2CMA analysis of cellular LRV1 and total RNA led to an estimate of before isolation of clonal lines. First drug treatment for 6.4 cell doublings about 15, consistent with estimates of LRV1 abundance from generates a population containing low average LRV1 levels, then the washout for 6 cell doublings allows resolution into fully negative or LRV1+ recent whole-genome RNA sequencing by our group. If this lines. (B) Representative LRV1 capsid profiles for a cured line (L. guyanensis unexpectedly low value for LRV1 copy number applies generally clone 10-5), a WT-like line (L. guyanensis clone 10-10), and a mixed profile to LRV1s in other Leishmania strains or species, it could provide line (L. guyanensis clone 10-1; for clarity the leading “10” is omitted from the figures). (C) RT-PCR tests confirming presence or absence of LRV1 in treated lines. RT+, reverse transcription performed before PCR; RT−,nore- verse transcription step. M, 1 kb+ ladder, Invitrogen. The expected LRV1 capsid and β-tubulin amplicons of 496 and ∼450 nt were found. (D) Western blotting with anti-LgyLRV1 capsid antisera confirms absence of LRV1 in cured lines L. guyanensis 10-5 and 10-6. M, molecular weight marker. The arrowhead marks the position of the 95-kDa LRV1 capsid band.

activity against HCV, where they target the viral RDRP by chain termination (37, 38, 49). By molecular modeling of the L. guyanensis LRV1 RDRP domain against other RDRPs, such as HCV, we were able to generate a view of the active site including residues putatively binding to the nucleotide substrates (Fig. S4). Notably, these included sites homologous to those mutated in HCV nucleoside analog-resistant lines (50). This finding supports our working hypothesis that both anti-LRV1 compounds are activated to triphosphates, where they act to inhibit RDRP activity. These compounds represent the only inhibitors known to act against any member of the Totiviridae, and indeed some of the few candidates described inhibiting dsRNA viruses generally. Common features of the two selective anti-LRV1 compounds include the 2′-C methyl and the adenine base moieties, although 2′ + − C-methyl G and C were inactive against both Leishmania and Fig. 7. Matched 2CMA-treated LRV1 and LRV1 cured lines recapitulate LRV1. A similar pattern was observed for dengue virus RDRP LRV1-dependent virulence. (A and B) Cytokine secretion by BMM infected 24 h inhibitors, where only adenosine analogs demonstrated antiviral after infection with L. guyanensis lines or treatment with poly I:C (2 μg/mL), activity (51). Following uptake, in Leishmania most purine nucleo- M, media. (A) TNF-α;(B) IL-6. The figure shown is representative of three sides are metabolized to nucleobases, the major exception being experiments, each done in triplicate; error bars represent ± SD. (C and D) μ + − adenosine, which is phosphorylated directly by adenosine kinase Infections of matched 10 M 2CMA treated LgyLRV1 and LgyLRV1 . Para- site numbers (luminescence from luciferase reporter) (C) or footpad swelling (31). This finding could contribute to the superiority of 2CMA (D) was measured at the peak of the infection (28 d). Each bar represents ′ + analogs. However, all other 2 C-modified analogs tested failed to pooled data from eight mice total, four for each Lgy line used. LRV1 (clones − inhibit LRV1 or Leishmania, including ones bearing adenine or 10-9 and 10-10) and LRV1 (clones 10-5 and 10-6) lines are shown; error bars related moieties as the nucleobase (Fig. 2 and Tables S1 and S2). represent ± SD. Data for control parasites are replotted from Ives et al. (7).

6of9 | www.pnas.org/cgi/doi/10.1073/pnas.1619114114 Kuhlmann et al. Downloaded by guest on September 25, 2021 a new perspective on the observation that thus far, no images of just yet, as the concentration needed for LRV1 elimination (10 μM) PNAS PLUS LRV1 in situ by electron microscopy appear in the literature. is above the maximum achievable serum concentration in various The rapid decline of LRV1 following 2CMA treatment sug- mammalian models, typically less than 1 μM (38, 49, 56). Thus, gested that it would be relatively easy to recover LRV1-free future efforts must focus on the development of compounds with clonal lines. Following washout of 2CMA after three to six cell higher potency targeting LRV1, without significant human host doublings and a brief period of growth without drug, cultures toxicity. For therapeutic purposes a compound simultaneously tar- manifested two distinct parasite populations by capsid flow geting both would likely be superior. However, because Leishmania + cytometry: one similar to LgyLRV1 and a second similar to growth is required for LRV1 to be lost by progressive dilution − − LgyLRV1 (Figs. 5 and 6). The fraction of LgyLRV1 parasites (Fig. 5), a screening method different from that used here will be grew progressively with increasing 2CMA treatment, reaching required to detect such agents. Dilutional loss following anti-LRV1 levels approaching 100%. To recover parasites suited for studies inhibitor treatment in vitro predicts that very low levels of LRV1 − focusing on the biological properties of LRV1 parasites, we could persist after treatment in vivo, whether measured on a total or adopted a protocol in which parasites were treated for only a per cell basis (Fig. 5). Importantly, previous data show that below a brief period with 10 μM 2CMA, a concentration showing little certain threshold, parasites bearing low LRV1 levels are controlled − effect on parasite growth but relatively high inhibition of LRV1 as effectively as LRV1 lines (7). (Fig. 3), followed by brief passaging and then plating on drug- Our studies also raise the possibility of treating other diseases free media. Importantly, this procedure allowed the recovery of caused by protozoans bearing dsRNA viruses, which show endoge- + − both LRV1 and LRV1 matched clonal lines, which had ex- nous virus-dependent pathogenicity, including Totiviridae present perienced identical treatment, thereby facilitating comparisons within Trichomonas vaginalis (Trichomonasvirus), Giardia lamblia probing LRV1 effects (below). Interestingly, in all of these (Giardiavirus), or Eimeria (Eimeravirus) (34, 57), and Partitiviridae studies the LRV1 levels in washout lines showed a strong ten- within Cryptosporidium parvum (Cryspovirus) (58, 59). Potentially, dency to recover from the low levels seen in drug to those agents targeting these putative pathogenicity factor viruses could + comparable to LRV1 controls (Fig. 5). These findings suggest prove similarly valuable for laboratory studies of these viruses as well. that the LRV1 copy number is maintained at a specific set point, perhaps through a balance between replication and the RNAi Methods pathway (12, 54). Previous studies examining LRV1 transcripts Parasites and Growth Media. Most studies were performed using luciferase- during growth phase also concluded that LRV1 copy number is expressing transfectants of L. guyanensis (MHOM/BR/78/M4147) described + −

previously [LRV1 LgyM4147/SSU:IR2SAT-LUC(b)c3 and LRV LgyM4147/ MICROBIOLOGY regulated (48). + pX63HYG/SSU:IR2SAT-LUC(b)c4] (54); these lines are termed LgyLRV1 and For other fungal dsRNA viruses, treatments engendering cell − + stress or growth inhibition have been used to generate virus-free LgyLRV1 , respectively. Two strains of LRV1 L. braziliensis were examined: LEM2780 (MHOM/BO/90/CS) and LEM3874 (MHOM/BO/99/IMT252 no. 3) (12). lines at significant frequencies, one common example being the Parasites were grown in Schneider’s media (Sigma) prepared according to use of CHX to cure the yeast L-A virus (44). Although in one the supplier’s instructions with pH adjusted to 6.5 and supplemented with prior study LRV1 cure was obtained during a series of trans- 0.76 mM hemin, 2 μg/mL biopterin, 50 U/mL penicillin, 50 μg/mL strepto- fection and hygromycin selection steps, this appears to have been mycin, and 10% (vol/vol) heat-inactivated FBS. Cell concentrations were successful only once, and our laboratories have been unable to determined using a Coulter Counter (Becton Dickinson). repeat this (12, 46). Here we were unable to show any correlation between LRV1 loss and drug-induced stress or growth inhibition Drug-Inhibition Tests. Compounds were purchased or obtained as summarized with CHX, the ergosterol synthesis inhibitor clotrimazole, or in Table S1, and the structures of the two most active anti-LRV1 compounds within the large panel of test compounds (Figs. 2 and 4, Fig. S2, are shown in Fig. S1. Stock solutions were prepared as recommended by the μ and Tables S1 and S2). Thus, LRV1 appears to be relatively source, typically in DMSO at 50 mM, and tested against parasites at 100 M or the maximum concentration permitted by drug solubility (Table S2). stable to growth inhibitory stresses. However, given its relatively 5 Parasites were inoculated at 2 × 10 cells/mL into Schneider’s media lacking low cellular copy number (<20), on a strictly probabilistic basis − supplemental adenine. Growth was evaluated after 2 d, before the controls LRV1 variants might occur at a low frequency, which occa- reached stationary phase growth, at which time parasite numbers had in- sionally may emerge or be recovered by methods more sensitive creased nearly 100-fold. Experiments were performed in sets of 10 test + than used here. compounds, along with LRV1 and negative controls; the agreement among independent experiments among the controls was excellent, and the results − Antiviral Cures and the Generation of Isogenic LRV1 Lines for the Study of are shown averaged together across all experiments (Table S2). LRV1-Dependent Virulence. Treatment with 2CMA enables the con- + trolled and reproducible generation of matched LRV1 and LRV1- LRV1 Capsid Flow Cytometry. For capsid flow cytometry, 107 cells were fixed at − cured lines without difficulty. In vivo, 2CMA-cured LRV1 parasites room temperature using 2% (wt/vol) paraformaldehyde (Thermo Fisher) in showed less pathology and lower parasite numbers and induced less PBS for 2 min. They were then incubated in blocking buffer [10% (vol/vol) + cytokine secretion than LRV1 parasites, comparable to the single normal goat serum (Vector Laboratories) and 0.2% Triton X-100 in PBS] for − 30 min at room temperature. Anti-LgyLRV1 capsid antibody (35) was added spontaneous LRV1 lines described previously (Fig. 7). Thus, our (1:20,000 dilution) and incubated at room temperature for 1 h. After two LRV1 toolkit now includes two independent, reproducible, and ef- μ − washes with PBS, cells were resuspended in in 200 L PBS with Alexa ficient methods for generating isogenic LRV1 lines, which will fa- Fluor488-labeled goat anti-rabbit IgG (Alexafluor, Invitrogen; 1:1,000; or cilitate tests probing the biology of LRV1-dependent pathogenicity in Thermo Fisher; 1:2,000 dilution), and incubated 1 h at room temperature. diverse parasite backgrounds. Depending on the relative selectivity of After two additional washes with PBS, cells were subjected to flow cytom- the antivirals and the presence of an RNAi pathway, one method etry, gating for single cells using forward and side scatter and the data may be superior for a given Leishmania speciesorstrain. analyzed using CellQuest software (BD Bioscience).

7 The Potential for Antitotiviral Therapy in the Treatment of dsRNA- RNA Purification, cDNA Preparation, and qRT-PCR. For RNA purification, 10 μ Bearing Parasites and Disease. There are now ample data sug- cells were resuspended in 350 L TRIzol Reagent and RNA was extracted gesting that LRV1 contributes to the severity in human leish- using the Direct-zol RNA purification kit according to protocol (Zymo Re- search). RNA was then treated with DNase I (Ambion) for 1 h at 37 °C and maniasis (6, 13, 17, 19, 20, 55), suggesting that anti-LRV1 repurified using RCC-5 column purification (Zymo Research). cDNA was inhibitors could be clinically useful, alone or in conjunction with prepared using SuperScript III (Invitrogen) and random priming according to existing antileishmanials. Unfortunately, pharmacokinetic stud- protocol. RNA denaturation occurred at 65 °C for 5 min. RT-PCR tests were ies of the two compounds studied here in mammals suggest that performed using LRV1-specific primers (SMB4647 5′-TBRTWGCRCACAGTGAY- neither of these are good candidates for testing of this hypothesis GAAGG and SMB4648 5′CWACCCARWACCABGGBGCCAT) or β-tubulin mRNA

Kuhlmann et al. PNAS Early Edition | 7of9 Downloaded by guest on September 25, 2021 (SMB5023 5′-AACGCTATATAAGTATCAGTTTCTGTACTTTA and SMB2110 vogen and used at 2 μg/mL. For mouse infections, 5- to 6-wk-old C57BL/6 5′-GACAGATCTCATCAAGCACGGAGTCGATCAGC). qRT-PCR was performed as mice were purchased from Jackson Laboratories. Parasites were grown into previously described (36), with a 123-bp fragment of LRV1 capsid mRNA am- stationary phase (2 full days) and 106 parasites were injected on the plantar ′ plified with primers SMB5335 (5 -CTGACTGGACGGGGGGTAAT) and SMB5336 aspect of the left foot. Measurement of footpad swelling was performed ′ (5 -CAAAACACTCCCTTACGC), and a 100-bp fragment of KMP-11 (a Leishmania weekly using a Vernier caliper. Parasite numbers were assessed by lumi- ′ housekeeping gene) with primers SMB5548 (5 -GCCTGGATGAGGAGTTCAACA) nescence of an integrated firefly luciferase reporter, measured using an IVIS and SMB5549 (5′-GTGCTCCTTCATCTCGGG). The reaction used Power SYBR 100 instrument as described previously (7, 54) and analyzed with Living Green (Applied Biosystems) in an ABI Prism 7000. Initial denaturation was at 95 °C Image software v2.60 (Perkin-Elmer). for 10 min followed by 40 cycles of amplification with 15 s at 95 °C, and 1 min at 60 °C. Data were analyzed using ABI 7000 SDS software (v1.2.3) and normalized using the ΔΔCT method (60). RNA slot blot analysis was performed as described Statement Identifying Institutional and Licensing Committee Approving Animal previously (36). The LRV1 copy number per cell was estimated in comparison with Experiments. Animal handling and experimental procedures were undertaken a standard curve generated using DNA from a plasmid bearing the LRV1 capsid with strict adherence to ethical guidelines relevant in both host countries. These gene (B6760, pBSKLRV1-4) and the average yield of RNA per cell across multiple are set out by the Swiss Federal Veterinary Office and under inspection by the L. guyanensis RNA preparations (5.12 ± 1.17 μg/107 cells; n = 34). Department of Security and Environment of the State of Vaud, Switzerland. Experiments were carried out in strict accordance with the recommendations in + − Isolation of LRV1 and LRV1 Clonal Lines by Brief Treatment with 2CMA. the GuidefortheCareandUseofLaboratoryAnimalsof the National Insti- + LgyLRV1 parasites were grown for one passage in media containing tutes of Health (61). Animal studies were approved by the Animal Studies 25 μg/mL nourseothricin (Werner BioAgent) to verify the presence of the Committee at Washington University (protocol #20090086) in accordance with integrated luciferase (LUC) gene (54). Cells were then grown one passage in the Office of Laboratory Animal Welfare’s guidelines and the Association for ’ the absence of nourseothricin, and inoculated into Schneider s media at a Assessment and Accreditation of Laboratory Animal Care International. concentration of 2 × 105 cells/mL into media containing 10 μM 2CMA. Growth was measured and LRV1 quantitated by capsid flow cytometry. At ACKNOWLEDGMENTS. We thank N. S. Akopyants, E. A. Brettmann, D. E. Dobson, various times, cells were either plated directly, or transferred to drug-free L.-F. Lye, and S. Schlesinger for discussions and comments on this manuscript; media, and allowed to grow an additional six cell doublings before plating. Chantal Desponds, Florence Prevel, and Haroun Zangger for excellent technical μ For both, the semisolid M199 media contained 50 g/mL nourseothricin, and assistance; and Jean Patterson (Texas Biomedical Research Institute) for pro- ∼ cells were diluted so that no more than 100 colonies were obtained per plate. viding capsid antisera. This work was supported by NIH Grants R01AI029646 and R56AI099364 (to S.M.B.); Fonds National de la Recherche Scientifique Macrophage Infections, Cytokine Assays, and Mouse Infection. Infections of Grants 3100A0-116665/1 and IZRJZ3_164176/1 (to N.F.); Sigma-Aldrich Pre- C57BL/6 mouse bone marrow-derived macrophages and cytokine assays were doctoral and the Sondra Schlesinger Graduate Student fellowships (to J.I.R.); performed as previously described (7, 10). Poly I:C was obtained from Invi- and the Division of Infectious Diseases (F.M.K.).

1. Alvar J, et al.; WHO Leishmaniasis Control Team (2012) Leishmaniasis worldwide and 19. Cantanhêde LM, et al. (2015) Further evidence of an association between the pres- global estimates of its incidence. PLoS One 7(5):e35671. ence of Leishmania RNA virus 1 and the mucosal manifestations in tegumentary 2. Bañuls AL, et al. (2011) Clinical pleiomorphism in human leishmaniases, with special leishmaniasis patients. PLoS Negl Trop Dis 9(9):e0004079. mention of asymptomatic infection. Clin Microbiol Infect 17(10):1451–1461. 20. Bourreau E, et al. (2016) Presence of Leishmania RNA virus 1 in Leishmania guyanensis 3. Singh OP, Hasker E, Sacks D, Boelaert M, Sundar S (2014) Asymptomatic Leishmania increases the risk of first-line treatment failure and symptomatic relapse. J Infect Dis infection: A new challenge for Leishmania control. Clin Infect Dis 58(10):1424–1429. 213(1):105–111. 4. Pigott DM, et al. (2014) Global distribution maps of the leishmaniases. eLife 3:3. 21. Castellucci LC, et al. (2014) Host genetic factors in American cutaneous leishmaniasis: 5. WHO (2010) Control of the leishmaniases: WHO Expert Committee on the Control A critical appraisal of studies conducted in an endemic area of Brazil. Mem Inst of Leishmaniases. World Health Organization Technical Report Series 949 (WHO, Oswaldo Cruz 109(3):279–288. Geneva). 22. Schriefer A, Wilson ME, Carvalho EM (2008) Recent developments leading toward a 6. Hartley MA, Drexler S, Ronet C, Beverley SM, Fasel N (2014) The immunological, paradigm switch in the diagnostic and therapeutic approach to human leishmaniasis. environmental, and phylogenetic perpetrators of metastatic leishmaniasis. Trends Curr Opin Infect Dis 21(5):483–488. Parasitol 30(8):412–422. 23. Parmentier L, et al. (2016) Severe cutaneous leishmaniasis in a human immunodefi- 7. Ives A, et al. (2011) Leishmania RNA virus controls the severity of mucocutaneous ciency virus patient coinfected with Leishmania braziliensis and its endosymbiotic leishmaniasis. Science 331(6018):775–778. virus. Am J Trop Med Hyg 94(4):840–843. 8. Widmer G, Comeau AM, Furlong DB, Wirth DF, Patterson JL (1989) Characterization 24. Eren RO, et al. (2016) Mammalian innate immune response to a Leishmania-resident of a RNA virus from the parasite Leishmania. Proc Natl Acad Sci USA 86(15): RNA virus increases macrophage survival to promote parasite persistence. Cell Host – 5979–5982. Microbe 20(3):318 328. 9. Stuart KD, Weeks R, Guilbride L, Myler PJ (1992) Molecular organization of Leish- 25. Castiglioni P, et al. (2016) Exacerbated leishmaniasis caused by a viral endosymbiont mania RNA virus 1. Proc Natl Acad Sci USA 89(18):8596–8600. can be prevented by immunization with its viral capsid. PLoS Negl Trop Dis, in press. 10. Zangger H, et al. (2014) Leishmania aethiopica field isolates bearing an endo- 26. De Clercq E, Li G (2016) Approved antiviral drugs over the past 50 years. Clin Microbiol – symbiontic dsRNA virus induce pro-inflammatory cytokine response. PLoS Negl Trop Rev 29(3):695 747. 27. La Frazia S, et al. (2013) Thiazolides, a new class of antiviral agents effective against Dis 8(4):e2836. 11. Scheffter SM, Ro YT, Chung IK, Patterson JL (1995) The complete sequence of Leish- rotavirus infection, target viral morphogenesis, inhibiting viroplasm formation. J Virol 87(20):11096–11106. mania RNA virus LRV2-1, a virus of an Old World parasite strain. Virology 212(1): 28. King AMQ, Lefkowitz E, Adams MJ, Carstens EB (2011) Virus Taxonomy: Ninth Report 84–90. of the International Committee on Taxonomy of Viruses (Elsevier, Amsterdam). 12. Brettmann EA, et al. (2016) Tilting the balance between RNA interference and rep- 29. Fichorova RN, et al. (2012) Endobiont viruses sensed by the human host—Beyond lication eradicates Leishmania RNA virus 1 and mitigates the inflammatory response. conventional antiparasitic therapy. PLoS One 7(11):e48418. Proc Natl Acad Sci USA 113(43):11998–12005. 30. Clatworthy AE, Pierson E, Hung DT (2007) Targeting virulence: A new paradigm for 13. Hartley MA, et al. (2016) Leishmaniavirus-dependent metastatic leishmaniasis is pre- antimicrobial therapy. Nat Chem Biol 3(9):541–548. vented by blocking IL-17A. PLoS Pathog 12(9):e1005852. 31. Carter NS, Yates P, Arendt CS, Boitz JM, Ullman B (2008) Purine and pyrimidine me- 14. Guerra JA, et al. (2011) Mucosal leishmaniasis caused by Leishmania (Viannia) bra- tabolism in Leishmania. Adv Exp Med Biol 625:141–154. ziliensis and Leishmania (Viannia) guyanensis in the Brazilian Amazon. PLoS Negl 32. Pfaller MA, Marr JJ (1974) Antileishmanial effect of allopurinol. Antimicrob Agents Trop Dis 5(3):e980. Chemother 5(5):469–472. 15. Goto H, Lindoso JAL (2010) Current diagnosis and treatment of cutaneous and mu- 33. Carrion R, Jr, Ro YT, Patterson JL (2008) Leishmaniaviruses. Encyclopedia of Virology, – cocutaneous leishmaniasis. Expert Rev Anti Infect Ther 8(4):419 433. eds Mahy BWJ, Van Regenmortel MHV (Academic, Amsterdam), pp 220–224. 16. Mears ER, Modabber F, Don R, Johnson GE (2015) A review: The current in vivo 34. Wickner RB, Ghabrial SA, Nibert ML, Patterson JL, Wang CC (2011) Totiviridae. Virus models for the discovery and utility of new anti-leishmanial drugs targeting cuta- Taxonomy: Ninth Report of the International Committee on Taxonomy of Viruses, eds neous leishmaniasis. PLoS Negl Trop Dis 9(9):e0003889. King AMQ, Lefkowitz E, Adams MJ, Carstens EB (Elsevier, Amsterdam), p 650. 17. Adaui V, et al. (2016) Association of the endobiont double-stranded RNA virus LRV1 35. Cadd TL, Keenan MC, Patterson JL (1993) Detection of Leishmania RNA virus 1 pro- with treatment failure for human leishmaniasis caused by Leishmania braziliensis in teins. J Virol 67(9):5647–5650. Peru and Bolivia. J Infect Dis 213(1):112–121. 36. Zangger H, et al. (2013) Detection of Leishmania RNA virus in Leishmania parasites. 18. Pereira LdeO, et al. (2013) Severity of tegumentary leishmaniasis is not exclusively PLoS Negl Trop Dis 7(1):e2006. associated with Leishmania RNA virus 1 infection in Brazil. Mem Inst Oswaldo Cruz 37. Carroll SS, et al. (2003) Inhibition of hepatitis C virus RNA replication by 2′-modified 108(5):665–667. nucleoside analogs. J Biol Chem 278(14):11979–11984.

8of9 | www.pnas.org/cgi/doi/10.1073/pnas.1619114114 Kuhlmann et al. Downloaded by guest on September 25, 2021 38. Olsen DB, et al. (2004) A 7-deaza-adenosine analog is a potent and selective inhibitor 60. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real- PNAS PLUS of hepatitis C virus replication with excellent pharmacokinetic properties. Antimicrob time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25(4):402–408. Agents Chemother 48(10):3944–3953. 61. Committee on Care and Use of Laboratory Animals (2011) Guide for the Care and Use 39. Eltahla AA, Luciani F, White PA, Lloyd AR, Bull RA (2015) Inhibitors of the hepatitis C of Laboratory Animals (National Academies Press, Washington, DC), 8th Ed. virus polymerase; mode of action and resistance. Viruses 7(10):5206–5224. 62. Appleby TC, et al. (2015) Viral replication. Structural basis for RNA replication by the 40. Sofia MJ, Chang W, Furman PA, Mosley RT, Ross BS (2012) Nucleoside, nucleotide, and hepatitis C virus polymerase. Science 347(6223):771–775. non-nucleoside inhibitors of hepatitis C virus NS5B RNA-dependent RNA-polymerase. 63. Meng EC, Pettersen EF, Couch GS, Huang CC, Ferrin TE (2006) Tools for integrated J Med Chem 55(6):2481–2531. sequence-structure analysis with UCSF Chimera. BMC Bioinformatics 7:339. 41. Yin Z, et al. (2009) An adenosine nucleoside inhibitor of dengue virus. Proc Natl Acad 64. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJ (2015) The Phyre2 web portal Sci USA 106(48):20435–20439. for protein modeling, prediction and analysis. Nat Protoc 10(6):845–858. 42. Carrion R, Jr, Ro YT, Patterson JL (2003) Purification, identification, and biochemical 65. Wass MN, Kelley LA, Sternberg MJ (2010) 3DLigandSite: Predicting ligand-binding characterization of a host-encoded cysteine protease that cleaves a leishmaniavirus sites using similar structures. Nucleic Acids Res 38(Web Server issue):W469–473. gag-pol polyprotein. J Virol 77(19):10448–10455. 66. Lam AM, et al. (2014) Molecular and structural basis for the roles of hepatitis C virus 43. VanDusen WJ, et al. (1997) Adenine quantitation in yeast extracts and fermentation polymerase NS5B amino acids 15, 223, and 321 in viral replication and drug resistance. – media and its relationship to protein expression and cell growth in adenine auxo- Antimicrob Agents Chemother 58(11):6861 6869. trophs of Saccharomyces cerevisiae. Biotechnol Prog 13(1):1–7. 67. Qing J, et al. (2016) Resistance analysis and characterization of NITD008 as an – 44. Fink GR, Styles CA (1972) Curing of a killer factor in Saccharomyces cerevisiae. Proc adenosine analog inhibitor against hepatitis C virus. Antiviral Res 126:43 54. Natl Acad Sci USA 69(10):2846–2849. 68. Dunbrack RL, Jr (2002) Rotamer libraries in the 21st century. Curr Opin Struct Biol – 45. Bhatti MF, et al. (2011) The effects of dsRNA mycoviruses on growth and murine 12(4):431 440. virulence of Aspergillus fumigatus. Fungal Genet Biol 48(11):1071–1075. 69. Leyssen P, De Clercq E, Neyts J (2008) Molecular strategies to inhibit the replication of – 46. Ro YT, Scheffter SM, Patterson JL (1997) Hygromycin B resistance mediates elimina- RNA viruses. Antiviral Res 78(1):9 25. 70. Smith DB, et al. (2007) Design, synthesis, and antiviral properties of 4′-substituted tion of Leishmania virus from persistently infected parasites. J Virol 71(12):8991–8998. ribonucleosides as inhibitors of hepatitis C virus replication: the discovery of R1479. 47. Wickner RB, Fujimura T, Esteban R (2013) Viruses and prions of Saccharomyces cer- Bioorg Med Chem Lett 17(9):2570–2576. evisiae. Adv Virus Res 86:1–36. 71. Tuttle JV, Tisdale M, Krenitsky TA (1993) Purine 2′-deoxy-2′-fluororibosides as anti- 48. Chung IK, et al. (1998) Generation of the short RNA transcript in Leishmaniavirus influenza virus agents. J Med Chem 36(1):119–125. correlates with the growth of its parasite host, Leishmania. Mol Cells 8(1):54–61. 72. Smith RA, Sidwell RW, Robins RK (1980) Antiviral mechanisms of action. Annu Rev 49. Eldrup AB, et al. (2004) Structure-activity relationship of heterobase-modified 2′-C-methyl Pharmacol Toxicol 20:259–284. ribonucleosides as inhibitors of hepatitis C virus RNA replication. J Med Chem 47(21): 73. Thio CL (2009) Hepatitis B and human immunodeficiency virus coinfection. Hepatology 5284–5297. 49(5, Suppl):S138–S145. 50. Migliaccio G, et al. (2003) Characterization of resistance to non-obligate chain-ter- 74. Razonable RR (2011) Antiviral drugs for viruses other than human immunodeficiency minating ribonucleoside analogs that inhibit hepatitis C virus replication in vitro. virus. Mayo Clin Proc 86(10):1009–1026. J Biol Chem 278(49):49164–49170. 75. De Clercq E, Holý A (2005) Acyclic nucleoside phosphonates: A key class of antiviral 51. Chen YL, et al. (2010) Inhibition of dengue virus RNA synthesis by an adenosine nu- drugs. Nat Rev Drug Discov 4(11):928–940. cleoside. Antimicrob Agents Chemother 54(7):2932–2939. 76. De Clercq E (2010) Antiretroviral drugs. Curr Opin Pharmacol 10(5):507–515. MICROBIOLOGY 52. Furman PA, Lam AM, Murakami E (2009) Nucleoside analog inhibitors of hepatitis C 77. Walker RT, DeClercq E, Eckstein F (1979) Nucleoside Analogues Chemistry, Biology, viral replication: Recent advances, challenges and trends. Future Med Chem 1(8): and Medical Applications (Plenum Press, New York). – 1429 1452. 78. Kitchin JE, Pomeranz MK, Pak G, Washenik K, Shupack JL (1997) Rediscovering my- ′ ′ ′ 53. Murakami E, et al. (2008) The mechanism of action of beta-D-2 -deoxy-2 -fluoro-2 - cophenolic acid: A review of its mechanism, side effects, and potential uses. JAm ′ ′ C-methylcytidine involves a second metabolic pathway leading to beta-D-2 -deoxy-2 - Acad Dermatol 37(3 Pt 1):445–449. ′ ′ fluoro-2 -C-methyluridine 5 -triphosphate, a potent inhibitor of the hepatitis C virus 79. Vaughan MH, Steele MW (1971) Differential sensitivity of human normal and ma- – RNA-dependent RNA polymerase. Antimicrob Agents Chemother 52(2):458 464. lignant cells to 8-azahypoxanthine in vitro. Exp Cell Res 69(1):92–96. 54. Lye L-F, et al. (2010) Retention and loss of RNA interference pathways in trypano- 80. Nelson JA, Carpenter JW, Rose LM, Adamson DJ (1975) Mechanisms of action of somatid protozoans. PLoS Pathog 6(10):e1001161. 6-thioguanine, 6-mercaptopurine, and 8-azaguanine. Cancer Res 35(10):2872–2878. 55. Ito MM, et al. (2015) Correlation between presence of Leishmania RNA virus 1 and 81. Shnider BI, et al. (1960) Clinical studies of 6-azauracil. Cancer Res 20:28–33. clinical characteristics of nasal mucosal leishmaniosis. Rev Bras Otorrinolaringol (Engl 82. Marr JJ, Berens RL (1983) Pyrazolopyrimidine metabolism in the pathogenic trypa- Ed) 81(5):533–540. nosomatidae. Mol Biochem Parasitol 7(4):339–356. 56. Carroll SS, et al. (2009) Robust antiviral efficacy upon administration of a nucleoside 83. Bluemling GR (2011) Part I: Synthesis of Cyclobutyl Nucleoside Analogs That Mimic analog to hepatitis C virus-infected chimpanzees. Antimicrob Agents Chemother AZT for Inhibition of the K65R HIV-1 Reverse Transcriptase Mutant. PhD thesis (Emory 53(3):926–934. University, Atlanta, GA). 57. Wu B, et al. (2016) Eimeria tenella: A novel dsRNA virus in E. tenella and its complete 84. Long PH (1941) The clinical use of sulfanilamide, sulfapyridine, sulfathiazole, sulfa- genome sequence analysis. Virus Genes 52(2):244–252. guanidine, and sulfadiazine in the prophylaxis and treatment of infections. Can Med 58. Nibert ML, Woods KM, Upton SJ, Ghabrial SA (2009) Cryspovirus: A new genus of Assoc J 44(3):217–227. protozoan viruses in the family Partitiviridae. Arch Virol 154(12):1959–1965. 85. Durantel D (2009) Celgosivir, an alpha-glucosidase I inhibitor for the potential 59. Ghabrial SA, et al. (2011) Partitiviridae. Virus Taxonomy: Ninth Report of the treatment of HCV infection. Curr Opin Investig Drugs 10(8):860–870. International Committee on Taxonomy of Viruses, eds King AMQ, Adams MJ, 86. Sawyer PR, Brogden RN, Pinder RM, Speight TM, Avery GS (1975) Clotrimazole: A Carstens EB, Lefkowitz EJ (Elsevier, Amsterdam), pp 523–534. review of its antifungal activity and therapeutic efficacy. Drugs 9(6):424–447.

Kuhlmann et al. PNAS Early Edition | 9of9 Downloaded by guest on September 25, 2021